the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Mar 2026
Scale-dependent spatial coherence between historical and instrumental earthquake catalogues at the global scale
Abstract. Historical earthquake catalogues extend seismic observations back by several centuries and are widely used in seismic hazard and tectonic studies, yet their global-scale informational content remains difficult to quantify due to strong spatial, temporal, and magnitude-dependent reporting biases. In this study, we present a quantitative, spatially explicit assessment of the consistency between global historical (1600–1899) and early instrumental (1900–1950, ) earthquake catalogues. Rather than relying on magnitude-based comparisons, we represent earthquake occurrence as spatial probability density fields obtained through Gaussian smoothing and define a scale-dependent spatial coherence metric based on the overlap between historical and instrumental distributions. This approach allows us to isolate large-scale tectonic signal from localized reporting artefacts and to systematically explore the role of spatial scale. Our results show that spatial coherence between historical and instrumental seismicity is low at small scales and increases monotonically with smoothing length, reaching moderate values only at regional to continental scales. Even at the largest scales considered, coherence remains well below unity, indicating that only a limited fraction of the instrumental spatial pattern is recoverable from historical data. Decomposition by tectonic domain reveals that subduction zones dominate the historical–instrumental agreement, while continental collision belts and intraplate regions contribute substantially less. These findings demonstrate that global historical earthquake catalogues contain a detectable but intrinsically limited imprint of tectonic structure. Meaningful use of historical seismicity at the global scale therefore requires explicit consideration of spatial scale and tectonic context. The framework proposed here provides a transparent and reproducible basis for evaluating the reliability of historical earthquake data in seismic hazard and global seismotectonic applications.
- Preprint
(1957 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-521', Anonymous Referee #1, 20 Apr 2026
-
RC2: 'Comment on egusphere-2026-521', Anonymous Referee #2, 20 Apr 2026
In the manuscript entitled ‘Scale-dependent spatial coherence between historical and instrumental earthquake catalogues at the global scale’ the author attempts to verify to what degree do global earthquake reports accord with the tectonic settings. In my opinion, the main strength of the manuscript is the decision to focus on the spatial distribution rather than on magnitude comparisons, thereby avoiding a well-known weakness of historical catalogues. This is a careful, consistent, and convincing methodological choice. I find the manuscript is well structured and interesting to read.
I think that in its present form, the manuscript lacks significant support and justification and could benefit from additional, deeper revision prior to publication. First and foremost, I am not sure the research questions (lines 78-83) are fully addressed. The thing that bothers me most is that the author attempts to resolve these goals solely by quantitative approaches whereas the nature of the pre-instrumental records is qualitative and should be inspected also is association with social/cultural/human context. The author is indeed aware of that (lines 192-193) but seems to ignore these important insights in his interpretation. Simply put, the fact that a given report exists is primarily subjected to human factors rather than tectonic settings. For example, large, populated regions most likely will be reported massively regardless of the type and nature of the nearby tectonic units. Such a discussion, rooted in the historical context of the data, is not present in the manuscript which limits the importance and meaningfulness of results and discredit the ability to accurately address the questions raised by the author.
The second point is related to the goal of the study: “with the aim of clarifying the extent to which historical seismicity reflects the underlying tectonic organization of earthquake occurrence” (lines 238-240). I find the conclusions of the manuscript regarding this aim is somewhat vague and in sufficient. For instance, to what extent do the reports reflect the underlying tectonic organization? How much? Where exactly? why? Figure 4 (c), 6 and 7 are valuable but the when the inspection is global (worldwide) one would expect a deeper evaluation of the results. While we are on that matter, I would argue that the coherence in Europe region (Figure 4(c)) demonstrates perfectly my previous point that is, the observed coherence is primarily subjected to human factors and NOT because of tectonic organization. I am not saying the author is incorrect – I only think that the conclusions are based on partial examination that lacks the historical context.
The third point is the geographic scales. I had difficulties in understand the scale definitions. The smoothing scale in km is fine and understandable, but I didn’t find any definitions for the geographic grid (line 115) as well as its characteristics (e.g., size, area etc) and why they were chosen (is it 200 km, line 159?). Additionally, throughout the manuscript the geographic definitions could be sharped. For instance, what is ‘fine spatial resolution (line 247) or regional scale (line 114)
Forth, the breakdown by tectonic domains is interesting, but the classification into only three domains and the use of a fixed distance of 200 km from plate boundaries seem simplistic (line 159). It would be worth justifying this choice or testing the sensitivity to alternative definitions of domains and distances of influence.
Fifth, in some sections and paragraphs, I find the phrasing to be vague and cumbersome (see or instance lines 188-194). Perhaps another round of edit will sharpen the content and meaning of the text.
Few more comments:
- Figure 1: Enlarge sub-figure (a) for it is the most important representation. If that map will be structured correctly, then the others might be omitted. Also emphasize the epicentres’ location – one can hardly detect some of them. What does the author mean by ‘All maps are shown using equal-area cartographic projections with coastlines for geographical reference’ (line 397-398)? Perhaps it would be better to describe the actual Coordinate Reference System and the datum used.
- Lines 418-419: “The resulting maps primarily reflect the evolving spatial heterogeneity of historical documentation and catalog compilation, rather than physical variations in seismicity”. See my first comment
- The use of Gaussian smoothing is good. I think ANNI (average nearest neighbor index) might add further statistical support/verification
- Lines 198-202 contain an internal contradiction which undermine the core claim of the manuscript. In the 1st sentence, the author suggests dominance of major tectonic structure while in the second he underscores the influence of reporting bias: “These features broadly correspond to major tectonic structures, such as subduction zones and active plate boundaries, suggesting that historical earthquake locations contain a non-random tectonic signal when considered at sufficiently large spatial scales. However, even at these scales, the density fields remain far from globally uniform, underscoring the persistent influence of reporting bias”
- Figure 3 might benefit from adding the polylines of the main tectonic units to emphasize the reporting densities superimposed on the tectonic settings.
- Figure 4: isn’t the results too obvious? The instrumental records clearly reflect the setting of the tectonic units since they are complete (or close to complete) while the historical records are partial and depend on human and historical context. What do we learn that we didn’t know from these results?
- Lines 207-208: ‘The global spatial coherence, quantified as the integrated overlap between the two density fields, remains substantially below unity, indicating that only a fraction of the instrumental spatial pattern is captured by historical reporting.’ What does the author mean by ‘below unity’?
- Lines 214-215: I partially agree with the statement. Yes, there are inconsistencies in historical magnitude estimations but that does not mean the data is ‘unsuitable for direct quantitative comparison’ but rather requires calibration and constraints within reasonable boundaries.
- Figure 5 should appear in the methodology section where the author explains why he does not rely on magnitudes.
- Lines 216-222: why this comparison was chosen and what do we learn from the results?
- Figures 6 and 7 could merge into one figure. I would also add visual representation on a map to the 3 inspected tectonic units
- The last paragraph of the results section contain inaccuracy: one can detect tectonic units from historical data also for local/regional scale (e.g., the DST or the NAF regions)
- The author is the sole contributor then why does he use the first-person plural(we) (e.g., line 241)
- Line 249-256: the author describes coherence of the different tectonic units, but it does not necessarily arise from the results. Please further explain.
- Repeating sentences and meanings: e.g., lines 256-259 were already stated in line 213
- Lines 261-262: I am not sure I agree with this statement. The author should further explain and justify. The same goes for lines 267-269
- Data and python code: should be but are not accessible
Citation: https://doi.org/10.5194/egusphere-2026-521-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 504 | 174 | 44 | 722 | 29 | 41 |
- HTML: 504
- PDF: 174
- XML: 44
- Total: 722
- BibTeX: 29
- EndNote: 41
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The paper investigated the spatial coherence between historical (1600-1899) and instrumental (1900-1950) catalogs, trying to answer three central questions (as out by the author):
i.
At which spatial scales does historical seismicity exhibit statistically meaningful agreement with
instrumental seismicity?
ii.
How does this coherence depend on tectonic setting (e.g. subduction zones, continental collision belts,
intraplate regions)?
iii.
What are the implications of these results for the use of global historical earthquake catalogs in seismic
hazard and tectonic studies?
To do so the author discussed features of the historical and instrumental catalogs and computes spatial probability density fields via isotropic Gaussian kernel smoothing. The author focuses on the event location rather than magnitude and provides insights on the limitations of the use of historical earthquakes catalogs for use in geophysical and hazards research.
The questions posed by the authors are certainly worth of investigation, but I feel that more work is desirable before seeing this work published in NHESS. Below I outline my general comments and the detailed comments are included in pop-up notes (usually associated to highlighted text) in the attached PDF.
1. The author needs to justify the use of the early instrumental catalog (1900-1950) instead of the recent one (e.g., 1970-2020): given the aims of his work, one wonders why not using a more reliable and complete instrumental catalog instead of the more uncertain and incomplete one from the first part of the last century;
2. The comparison between historical and instrumental catalogs strikes me as unfair. It is clear that the European historical catalog lists many small earthquakes that would not be included in an instrumental global catalogs with M>5.5 (addressing point 1 could benefit the aims of the author, as instrumental catalogs for recent decades would include smaller earthquakes);
3. It feels from several sentences in the text that one goal of the author is to "downgrade" (somewhat) the importance of the historical earthquake catalog in the use of hazard and geophysical research. We all know that any earthquake catalog has limitations and that as we go further back in time such limitations are exacerbated by the paucity of data, however I feel that the tone of the text should just reflect that as in its current state it unnecessarily depicts the historical record too negatively.